What data format does it need?

Ideally one row per customer per month with customer_id, month and mrr. It also accepts invoice lists or already-aggregated figures.

What if I lack gross margin or sales spend?

The affected metrics (LTV, CAC, payback) are marked "data needed" instead of being guessed. Everything else is still computed.

Will it tell me if my numbers are good?

It gives directional guidance and labels any range as approximate, since it varies by segment and stage. Under 10 customers, churn percentages are noisy.

Built hereUpdated 10 Jul 2026

SaaS metrics: MRR, churn, LTV

/finance--saas-metrics

/finance--saas-metrics is a Claude Code skill in the Business section. Computes MRR, ARR, churn, NRR, LTV and CAC from your subscription data, showing each formula and a plain-language reading.

Author's description

Compute SaaS metrics from revenue data: MRR, ARR, churn, NRR, LTV, CAC

Use it when

You have subscription or invoice exports and want your metrics
You need MRR, churn and NRR month by month
You want LTV, CAC and payback and know your margin and sales spend

Not for

Financial or investment advice
Exact benchmarking: it only gives approximate rules of thumb

What you get

A monthly table in the chat, a CSV with the full series and a Markdown report with formulas, assumptions and interpretation.

How to ask for it

/finance--saas-metrics calculate this month's MRR and churn
/finance--saas-metrics I need the LTV and CAC for our customers
/finance--saas-metrics compute SaaS metrics from this revenue CSV

Install

curl -fsSL https://raw.githubusercontent.com/sgomez-dev/claude-skills/main/install.sh | bash

After installing with the script, type /finance--saas-metrics. Using Cursor, Windsurf or Codex? Platform guides

Questions about this skill

What data format does it need?: Ideally one row per customer per month with customer_id, month and mrr. It also accepts invoice lists or already-aggregated figures.
What if I lack gross margin or sales spend?: The affected metrics (LTV, CAC, payback) are marked "data needed" instead of being guessed. Everything else is still computed.
Will it tell me if my numbers are good?: It gives directional guidance and labels any range as approximate, since it varies by segment and stage. Under 10 customers, churn percentages are noisy.

Demo coming soon